This paper by Sebastian Benthall frames it as disciplinary collapse, connecting the ease with which interdisciplinary interactions take place on social media to the broader phenomenon of context collapse:
This paper is the result of context collapse. Davis and Jurgenson review how the collapse of social contexts is a noted characteristic of interactions in networked publics, and distinguish between context collision, or accidental collapse, and context collusion, or intentional collapse. [42] As more of academic discussion moves to open access journals and more researchers encounter each other online, the mechanism of context collapse plays more of a role in the encounters between normally contextually separated academic disciplines. These collapses may be disciplinary collisions or disciplinary collusions.
https://commons.pacificu.edu/work/sc/63bcdb50-8c4e-4823-b34b-178602ea9ee0
This excellent essay on Google Scholar (where I found the reference) frames this as a more ambiguous phenomenon than Benthall does, suggesting that STEM authorship and citation conventions will tend to be privileged within systems characterised by disciplinary collapse:
Other effects on academia also need close attention. For instance, the way Google Scholar’s relevance algorithm works appears to be shifting the distribution of knowledge and ‘impact’ between disciplines. Because it does not pay attention to the specificities of different academic fields, it privileges physical sciences and computer science over social sciences and humanities in terms of search results and citation counts. This is a feature of how the technology functions in combination with the different citation and authorship conventions in STEM fields. Google Scholar therefore participates in a broader paradigm shift towards computer science appearing as the discipline around which all knowledge is organized. Citation count comparison across an ontologically flat scholarly network contributes intellectually to disciplinary collapse (Benthall, 2015), while establishing the primacy of computer engineering as a matter of sociotechnical fact.
It’s not clear to me exactly what the mechanism is here. The shorter length, longer co-authorship teams and higher citation counts in STEM disciplines lead to a ‘crowding out’ of social science papers? It’s an interesting suggestion but it’s almost certainly a more complex picture than this account implies, even if I find the suggestion intuitively plausible.
